1j) as compared with the fulleroids bearing at least one less
bulky Me group (1a, 1c, 1d and 1h).11 Here, we assumed
that the first ring-closure is more effectively controlled by
the enthalpy change (∆H) than by the entropy change (∆S)
because of the apparent structual similarity between the [5,6]
open fulleroid and [5,6] closed isomer.
results are also compatible with the radical process in view
of the captodative acceleration in the case of CF3- and
cyclopropyl-substituted 1f13 as well as the poor dependency
of rate on the solvent polarity (as ET14) for 1g at 90 °C; i.e.,
105k/s-1 ) 1.35 (toluene, ET ) 33.9), 2.03 (ODCB, 38.0),
and 4.25 (1,1,2,2-tetrachloroethane, 39.4). Absence of the
â-scission products of a cyclopropyl group of 1c-g may be
ascribed to the rapid recombination of the possible biradical
intermediate like in norcaradiene walk rearrangement.15
However, a question is raised as to why the far more
intrinsically potent aryl groups did not contribute to the
stabilization of radical intermediate as compared to the
cyclopropyl groups.16 This can be explained by considering
the geometrically controlled stereoelectronic effects of the
relevant substituents.
As to the latter [1,5] shift, we employed the corrected
∆Gq by subtracting the above ∆H from the ∆Gq derived
corr
from the observed k values (Table 2). When ∆Gq was
corr
plotted against ∆H, a fairly well correlated line with the slope
of 0.82 (including 1a) was obtained for the aryl-substituted
fulleroids (filled circle) as shown in Figure 1. This means
As depicted in Figure 2, the cyclopropyl group will
promote the homolytic-cleavage of the methano-bridged bond
Figure 2. (a) Favorable π-conjugation of 1g with the bisected
cyclopropyl group; (b)less favorable π-conjugation of 1i with the
hindered aryl group.
Figure 1. Plots of ∆Gq
fulleroids 1.
vs ∆H for the rearrangement of
corr
(C1-C2 or C1-C3) by π-resonating stabilization of the
generating spin center in the bisected conformation (a)
typically represented for 1g. By contrast, such an orbital
interaction is severely inhibited in case of the aryl-substituted
fulleroids due to the congested steric repulsion between the
peri-hydrogen and the fullerene cage (b). This geometrical
explanation is supported by the activation parameters for
thermal rearrangement of dicyclopropyl-substituted 1g. The
plot of ln k vs 1/T yielded the extrapolated ∆Hq (25.4 kcal
mol-1) and ∆Sq (-10.3 cal mol-1 T-1) at 170 °C. Compared
with the rearrangement of bis(p-anisyl)-substituted fulleroid
(22.9 kcal mol-1 and -24.2 cal mol-1 T-1, respectively),2d
1g takes advantage of the far more favorable ∆Sq, reflecting
the facile prealignment of the two cyclopropyl groups for
the bisected conformation (a).
that the aryl group exerts the promoting effects both on the
ring-closure and the [1,5] shift, although the efficiency is
slightly larger in the former process. On the other hand, the
cyclopropyl-substituted fulleroids gave the scattered plots
(open circle) well below the aryl line. The lower deviation
is increased with the increasing number of cyclopropyl group
(1a < 1d e 1c < 1g), demonstrating the additivity of
cyclopropyl contribution (vide infra). This is also the case
for the tolyl-substituted fulleroids (1h < 1i), although the
efficacy is considerably reduced. Accordingly, the substituent
effect of cyclopropyl group on the [1,5] shift is found to be
more effective than that of aryl group.
Mechanistically, the [1,5] shift is invoked to proceed via
radical species by Shevlin et al. on the basis of the capto-
dative effects in aryl-substituted fulleroids as well as the
dramatic rate enhancement by the prealigned π-resonating
substituents.2d It may be also possible that the present [1,5]
shift is similar to the norcaradiene-type rearrangement, so-
called “walk rearrangement”, in which the proposed diradical
species rapidly recombines with the stereoinversion.12 Our
However, at present, we cannot thoroughly rule out the
possibility that the homolytic bond cleavage takes place in
(12) Kless, A.; Nendel, M.; Wilsey, S.; Houk, K. N. J. Am. Chem. Soc.
1999, 121, 4524.
(13) (a) Creary, X.; M.-Mohammadi, M. E. J. Org. Chem. 1986, 51,
2664. (b) Creary, X.; Sky, A. F.; M.-Mohammadi, M. E. Tetrahedron Lett.
1988, 29, 6839.
(14) Reichardt, C. Chem. ReV. 1994, 94, 2319.
(15) Based on the caluculated â-scission rate constant (3.11 × 109 s-1
)
(10) Minimum energy of the possible monocyclopropyl-substituted [5,6]
closed isomer was not obtained as in the case of the unsubstituted [5,6]
closed isomer; see ref 2a. This may be one of the reasons for the very slow
or negligible rearrangement of these [5,6] open fulleroids.
at 150 °C for the comparable dimethyl cyclopropyl carbinyl radical (ref 6)
and the limits of detectability (< 2 %) by 1H NMR mesurements, the
lifetimes of biradical intermediate from 1c and 1d were estimated to be
shorter than 6.4 ps.
(11) Such a steric bulk effect is well recognized in the valence
equilibration between 7-substituted cycloheptatriene and norcaradiene:
Takahashi, K.; Takase, K.; Toda, H. Chem. Lett. 1981, 979.
(16) The cyclopropyl group is less effective in radical stabilization than
the phenyl group: Engel, P. S.; Nalepa, C. J.; Horsey, D. W.; Keys, D. E.;
Grow, R. T. J. Am. Chem. Soc. 1983, 105, 7102.
Org. Lett., Vol. 10, No. 2, 2008
295
Kitamura, Hiroshi
